In the fall of 2011 I started a series of blog posts on NoSQL databases. Since then I have developed a conference session on the topic which I presented at DevTeach 2012 in Vancouver in May. In developing the talk I spent a good deal of time trying to understand the concept of Map/Reduce, as it can be a challenging topic.
As described earlier in this series, NoSQL databases can be partitioned across many systems (or servers). This gives rise to an interesting problem - how to query across these multiple servers. This problem was solved by the concept of Map/Reduce algorithms. Map/Reduce is a multi step querying process, which takes a big task and breaks it down into multiple smaller tasks, and as with many of the NoSQL innovations it was first developed at Google.